Hardware acceleration for Unikernels - A status update of vAccel

Describes hardware acceleration for unikernels via the vAccel framework, decoupling applications from hardware-specific implementations so accelerated (e.g., ML inference) workloads can run inside unikernels, with frontend/backend driver design over QEMU/KVM and demos.